当前位置：文档库 › 2004-01 热重分析很好水热制备镁取代HA

2004-01 热重分析很好水热制备镁取代HA

Biomaterials25(2004)4647–4657

Preparation of magnesium-substituted hydroxyapatite powders by the mechanochemical–hydrothermal method Wojciech L.Suchanek a,1,Kullaiah Byrappa a,2,Pavel Shuk a,3,Richard E.Riman a,*,

Victor F.Janas b,Kevor S.TenHuisen b,4

a Department of Ceramic and Materials Engineering,Rutgers University,607Taylor Rd,Piscataway,NJ08854-8087,USA

b Johnson&Johnson,Center for Biomaterials and Advanced Technologies,a Division of Ethicon,Rt.22W,P.O.Box151,Somerville,

NJ08876-0151,USA

Received3June2003;accepted7September2003

Abstract

Magnesium-substituted hydroxyapatite(Mg-HAp)powders with different crystallinity levels were prepared at room temperature via a heterogeneous reaction between Mg(OH)2/Ca(OH)2powders and an(NH4)2HPO4solution using the mechanochemical–hydrothermal route.The as-prepared products contained unreacted Mg(OH)2and therefore had to undergo puri?cation in ammonium citrate aqueous solutions at room temperature.X-ray diffraction,infrared spectroscopy,thermogravimetric and chemical analyses were performed and it was determined that the puri?ed powders were phase-pure Mg-HAp containing0.24–28.4wt%of Mg.The concentration of Mg was slightly lower near the surface than in the bulk of the HAp crystals as indicated by X-ray photoelectron spectroscopy.Dynamic light scattering revealed that the median particle size of the room temperature Mg-HAp powders was in the range of102nm–1.2m m with a speci?c surface area between91and269m2/g.Scanning electron microscopy con?rmed that the Mg-HAp powders consisted of submicron agglomerates of nanosized crystals,less than E20nm.

Keywords:Hydroxyapatite;Magnesium;Nanoparticle;Crystallinity;Hydrothermal synthesis;Mechanochemical synthesis

1.Introduction

Hydroxyapatite[HAp,chemical formula Ca10(PO4)6-(OH)2]has attracted attention of researchers during the past30years as an implant material because of its excellent biocompatibility and bioactivity[1,2].It has also been extensively investigated for non-medical applications such as packing media for column chro-matography,gas sensors,catalysts,host material for lasers,and plant growth substrates.All properties of HAp,including bioactivity,biocompatibility,solubility, and adsorption properties can be tailored over a wide range by modifying the composition through ionic substitutions[1–4].

Magnesium(Mg)has been known as one of the cationic substitutes for calcium in the HAp lattice[3,4]. Usually,the incorporation of Mg in synthetic HAp is limited(maximum of about0.4wt%of Mg)unless other ions,such as carbonate or?uoride are simultaneously incorporated together with magnesium as paired sub-stitutions[4].Increasing concentration of Mg in HAp has the following effects on its properties:(i)decrease in crystallinity,(ii)increase in HPO42àincorporation,and (iii)increase in extent of dissolution[4].Magnesium ions also have been found to inhibit growth of the(001)face of HAp crystals[5].These substitution-property rela-tionships allowtailoring physicochemical properties of HAp by controlling the Mg substitution.

Mg is one of the main substitutes for calcium in biological apatites.Enamel,dentin and bone contain, respectively,0.44, 1.23,and0.72wt%of Mg[4];

*Corresponding author.Tel.:+1-732-445-4946;fax:+1-732-445-6264.

E-mail addresses:w suchanek@https://www.wendangku.net/doc/ba6494679.html,(W.L.Suchanek), riman@https://www.wendangku.net/doc/ba6494679.html,(R.E.Riman).

1Present address:Sawyer Research Products,Inc.,35400Lakeland Boulevard,Eastlake,Ohio44095,USA.

2Permanent address:Department of Geology,University of Mysore, P.B.No.21,Mysore570006,India.

3Present address:Rosemount Analytical Inc.,1201North Main Street,P.O.Box901,Orrville,OH44667-0901,USA.

4Present address:Stryker Orthopaedics,325Corporate Drive, Mahwan,NJ07430,USA.

therefore,the Mg-substituted HAp materials(denoted hereafter as Mg-HAp)are expected to have excellent biocompatibility and biological properties.Mg-HAp ceramics have been proposed for use in orthopedic and dental applications.Bon?eld and Gibson suggested coupled Mg(up to0.5wt%)and CO3(up to1wt%) substitution in HAp synthesized by precipitation from aqueous solutions as a rawmaterial for bone implants fabrication[6].Dolci et al.claimed use of HAp powder (crystal size ranging between0.5and200nm)containing up to25.4wt%of Mg for the production of prepara-tions for odontostomatologic applications[7]. Magnesium is closely associated with mineralization of calci?ed tissues[4]and indirectly in?uences mineral metabolism[8].Its role,how ever,has not been fully understood.It has been suggested that magnesium directly stimulates osteoblast proliferation with an effect compar-able to that of insulin(a known growth factor for osteoblast)[9].However,in vitro studies show that partial replacement of Ca w ith Mg in the HAp crystals(5w t%)in HAp–collagen composite sponges resulted in a decrease of the proliferation and activities of the osteoblast-like cells [10].Total Mg substitution in HAp had a toxic effect on bone cells and prevented formation of an extracellural matrix[10].However,in that study,the mineral phase in the HAp–collagen composites has not been characterized in suf?cient detail.In addition,amounts of Mg in these composites were much higher than in the natural bone tissue.Similarly controversial results were observed in studies on the effects of Mg supplementation on natural bone health,as reviewed by Martini[11].

Since the optimum amounts of Mg in arti?cial HAp ceramics can vary with different applications,a cap-ability to control the substitutional level of Mg in HAp over the widest possible range by controlling the synthesis procedure is of signi?cant interest.Mg-HAp powders have been prepared by precipitation and hydrolysis methods indicating limited replacement of Ca2+with Mg2+(up to0.3wt%)[12].Some other studies on precipitation of Mg-HAp powders also showed limited amounts of Mg(less than1%)in HAp [4,13,14].Mayer et al.precipitated HAp powders containing up to1.5wt%of Mg without simultaneous carbonate substitution[15].Golden and Ming managed to synthesize from aqueous solutions Mg-HAp powders with up to2wt%of Mg[16].Bigi et al.synthesized Mg-HAp powders with up to5wt%of Mg in HAp under hydrothermal conditions at120 C[17].Okazaki et al. substituted up to5wt%of Mg in HAp using precipita-tion method but with a total loss of crystallinity[18].It is worth mentioning that a total replacement of Ca2+ with Mg2+in HAp powders precipitated in solution at high pH has been reported[19,20].However,other researchers were not able to reproduce these results[21]. Recently,several papers regarding mechanochemical and mechanochemical–hydrothermal syntheses of HAp powders appeared in the literature[22–29].Mechan-ochemical powder synthesis is a solid-state synthesis method that takes advantage of the perturbation of surface-bonded species by pressure to enhance thermo-dynamic and kinetic reactions between solids[30]. Pressure can be applied via conventional milling equipment ranging from low-energy ball mills to high-energy stirred mills(e.g.attrition,planetary,or vibra-tory mills).In the mechanochemical–hydrothermal process,water actively participates in the synthesis by both dissolving one of the reacting powders as well as serves as a reactant to produce a highly crystalline product with precisely controlled chemical composition, which is not observed with the conventional mechan-ochemical method.The main advantages of the mechan-ochemical–hydrothermal synthesis of ceramic powders are simplicity and lowcost.In our earlier studies,w e have demonstrated that crystalline HAp powders with either stoichiometric composition or with up to12wt% of carbonate substitution could be prepared at room temperature from heterogeneous reactions between Ca(OH)2/CaCO3/Na2CO3powders and(NH4)2HPO4 solution via the mechanochemical–hydrothermal route [31,32].This processing is advantageous when com-pared to prior reported mechanochemical HAp synth-eses,as discussed in detail in Refs.[31,32].The mechanochemical–hydrothermal processing can pro-duce nanosized,stoichiometric,highly crystalline,and thermally stable pure HAp powders,while the prior mechanochemical works report non-stoichiometric, thermally unstable materials with very low crystallinity levels[31,32].In the present paper,we will introduce a mechanochemical–hydrothermal method for prepara-tion of nanocrystalline Mg-HAp powders with con-trolled magnesium substitution.

2.Experimental procedure

2.1.Mechanochemical–hydrothermal synthesis Targeted chemical compositions corresponded to the x values of0,0.5,1,2,3,4,5,6,and10in the simpli?ed chemical formula of Mg-HAp,which is Ca10àx Mg xePO4T6eOHT2:This chemical formula does not take into account either increasing HPO42àsubstitu-tion or formation of related lattice defects with increasing Mg content;therefore,it can only be used as a rough approximation.Ca(OH)2,Mg(OH)2,and solid(NH4)2HPO4(all analytical grade,Alfa Aesar, Ward Hill,MA)were used as reactants for synthesis of Mg-HAp.Their purity was con?rmed by X-ray diffrac-tion and thermogravimetry.The presence of water adsorbed on all reactants was measured by thermo-gravimetry and then used to adjust the reactant quantities to maintain the targeted stoichiometries.

W.L.Suchanek et al./Biomaterials25(2004)4647–4657 4648

First,suspensions containing powder mixtures of0–23.200g Ca(OH)2and 1.003–20.000g Mg(OH)2in 350ml of deionized water were prepared with a500ml glass beaker.Subsequently,25.892–29.410g of (NH4)2HPO4powder was slowly added to the same beaker at constant vigorous stirring using a magnetic stirrer for about10min.The(Ca+Mg)/P molar ratio in the starting slurry was1.67.Slurry pH was measured using a glass electrode connected to a pH-meter (Accumet s Model805MP,Fisher Scienti?c,Pittsburgh, PA)and calibrated with respect to a buffer solution (pH=10.00,Fisher Scienti?c,Pittsburgh,PA).The measured pH of the slurries was in the range of9.94–10.44.The mechanochemical–hydrothermal synthesis was performed by placing the slurry into a laboratory-scale mill(model MIC-0,NARA Machinery Co., Tokyo,Japan)equipped with a zirconia liner and zirconia ring grinding media.

The milling equipment was a multi-ring media mill and its grinding mechanism is different from conven-tional attrition,planetary,or vibratory mills[33].The mill consists of a central rotating stainless-steel shaft, which drives six stainless-steel sub-shafts(sleeve-lined with zirconia-toughened alumina)that are connected symmetrically to the central shaft.Each sub-shaft contains19stacked zirconia rings,which can rotate eccentrically around each sub-shaft.When the central shaft is rotating,the zirconia rings on the sub-shafts are moved by the centrifugal force radially outwards applying force on a ceramic liner which is mounted on an inner wall of the milling vessel.Solid particles,which are located between the rotating rings and the liner wall, are subsequently comminuted[33]. Mechanochemical–hydrothermal reaction of the slur-ry was carried out in air,initially at a rotation speed of 1500rpm for1h and then at800rpm for4h.Tempera-ture during the reaction was measured using a thermo-couple and ranged between29–33 C at1500rpm and 25–28 C at800rpm.Washing of the solid phase after the mechanochemical–hydrothermal synthesis was ac-complished by2–6cycles of shaking the solid with distilled water in2–6HDPE250ml bottles using a hand shaker machine(Model M37615,Barnstead/Thermolyne, Dubuque,Iowa)followed by centrifuging at4500rpm for 30min(Induction Drive Centrifuge,Model J2-21M, Beckman Instruments,Fullerton,CA).The washed solid phase was dried in an oven at70 C for24h(Isotemp s oven,model230G,Fisher Scienti?c,Pittsburgh,PA)and ground into powder.These powders were referred to as ‘‘as-prepared Mg-HAp’’in the entire paper.

2.2.Puri?cation of the as-prepared Mg-HAp powders The as-prepared Mg-HAp powders contained differ-ent fractions of unreacted Mg(OH)2.Therefore they were subjected to a puri?cation treatment using0.2m-ammonium citrate aqueous solutions.Each ammonium

citrate solution was prepared in a200ml glass beaker by

dissolving 3.843g of solid citric acid(reagent grade,

Aldrich,Milwaukee,WI)in100ml of the distilled water

and subsequently slowly adding ammonia solution

(reagent grade,Fisher Scienti?c,Pittsburgh,PA)to

yield a pH between8and10.One gram of the powder

containing Mg-HAp with unreacted Mg(OH)2was then

suspended in this solution.Dissolution of the Mg(OH)2

was accomplished under a vigorous stirring using a

magnetic stirred for12–24h.In the samples containing

larger quantities of unreacted Mg(OH)2(x=3,4,5in

the simpli?ed Mg-HAp formula),it was necessary to

repeat this procedure in order to completely remove the

Mg(OH)2phase.Washing of the solid phase after the

ammonium citrate treatment was accomplished by3

cycles of shaking the solid with distilled water in50ml

bottles using a hand shaker machine(Model M37615,

Barnstead/Thermolyne,Dubuque,Iowa)followed by

centrifuging at4500rpm for30min(Induction

Drive Centrifuge,Model J2-21M,Beckman Instru-

ments,Fullerton,CA).The washed Mg-HAp

powders were dried in an oven at70 C for24h

(Isotemp s oven,model230G,Fisher Scienti?c,Pitts-

burgh,PA).These powders are referred to in this paper

as‘‘puri?ed Mg-HAp’’.

2.3.Characterization of the materials

In order to check thermal stability of the puri?ed Mg-

HAp powders herein referred to‘‘heat-treated’’,a small

quantity of each puri?ed Mg-HAp powder was placed in

an alumina crucible and heat treated in air at900 C for

1h with a heating rate of10 C/min(rapid temperature

furnace,CM Inc.,Bloom?eld,NJ).The samples were

cooled together in the furnace(average cooling rate E6 C/min)and removed from the furnace after cooling down to room temperature.

X-ray diffraction characterization of all batches of as-

prepared,puri?ed,and heat-treated Mg-HAp powders

was performed using Ni?ltered Cu K a radiation.

Samples were analyzed over a2y range of10–70 at a

scan rate of2.4 /min,with a sampling interval of0.05

(XRD,Kristallo?ex D-500,Siemens Analytical X-ray

Instrument Inc.,Madison,WI).Crystallographic iden-

ti?cation of the synthesized phases was accomplished by

comparing the experimental XRD patterns to standards

compiled by the Joint Committee on Powder Diffraction

and Standards(JCPDS),which were card#09-0432

for HAp,#07-0239for Mg(OH)2,#20-0663for

NH4MgPO4áH2O,#11-0231for Ca4Mg5(PO4)6,#09-

0169for whitlockite,and#33-0297for b-Ca2P2O7.

Speci?c surface area of all batches of puri?ed Mg-

HAp powders was measured using the BET method

utilizing adsorption of N2gas(purity99.99%,Mathe-

son Co.,Bridgeport,NJ)at—196 C(Micromeritics

W.L.Suchanek et al./Biomaterials25(2004)4647–46574649

2375,Micromeritics,Norcross,GA).For this purpose, 0.10–0.30g of the Mg-HAp powder was outgased for2–4h at120 C.Particle size of the primary crystals was estimated from the nitrogen adsorption isotherms using the BET method to calculate equivalent spherical diameter,or BET particle diameter(d BET)from the following equation:d BET=6/(ráS w),where r is density and S w is the speci?c surface area.A density of3.156g/ cm3,which is a theoretical density of stoichiometric HAp,was used for all calculations.

Particle size distributions of all batches of puri?ed Mg-HAp powders were determined by dynamic light scattering at a wavelength of632.8nm(DLS,model DLS-700,Otsuka Electronics Co.,Osaka,Japan). Samples for the DLS measurements were prepared by dispersing small amounts of the HAp powder in ethanol (?ltered using a0.2m m?lter)followed by treatment in an ultrasonic bath for10min.After transferring to the sample holder,the suspensions were diluted again using ?ltered ethanol and ultrasonicated for3min.The measurement conditions included a sampling time of 80m s and100accumulations.A viscosity of1.19cP and refractive index of1.36were used for calculations. Size and degree of agglomeration of the synthesized particles in selected batches of puri?ed Mg-HAp were studied using?eld emission scanning electron micro-scope(FESEM,Model DSM962,Gemini,Carl Zeiss, Inc.,Thornwood,NY)at1.0–2.0kV with a working distance of2–5mm.In order to prepare a sample for the FESEM analysis,a small quantity of the powder was suspended in ethanol,placed in an ultrasonic bath for 10min and transferred onto graphite tape mounted on sample holders.No conductive coating was used. Infrared spectra of all batches of as-prepared and puri?ed Mg-HAp powders were obtained using an infrared Fourier-transform spectrometer(FTIR,model 1720-X,Perkin Elmer Co.,Norwalk,CT).For this purpose,each powder was mixed with KBr in the

proportion1

150(by weight)for15min and pressed into a

pellet using a hand press.Thermogravimetric analysis (model TGA-6,Perkin Elmer Co.,Norwalk,CT)was performed on all batches of as-prepared and selected batches of puri?ed Mg-HAp powders using a heating rate of5 /min up to a maximum temperature of950 C in?owing air atmosphere at a?ow rate of20ml/min. Chemical analysis for Ca,Mg,P,and NH4in the selected batches of puri?ed Mg-HAp powders was accomplished by X-ray?uorescence spectroscopy (XRF,Oneida Research Services,Inc.,Whitesboro, NY).The error of the measured Mg concentrations using this technique was estimated to be up to735%of the measured value.X-ray photoelectron spectroscopy (XPS,spectrometer model5700Lsci ESCA with a monochromatic aluminum X-ray source of350W, Physical Electronics,Eden Prairie,MN)of selected puri?ed Mg-HAp samples(x=0.0and2.0)was accom-plished as follows.The samples were deposited in small troughs in a tantalum plate and initially analyzed using low-resolution survey scans to determine which elements were present.High-resolution spectra were then taken to determine the concentration and binding energy of the elements detected in the survey scan.The quanti?cation of the elements was accomplished by using the atomic sensitivity factors for the used spectrometer.The approximate sampling depths for the carbon1s photo-electron were30and90(A,for the electron exit angles of 20 and80 ,respectively.A region with a0.8mm diameter was analyzed for each sample.

3.Results and discussion

3.1.As-prepared Mg-HAp powders

XRD patterns of all as-prepared Mg-HAp powders are shown in Fig.1.Clearly de?ned HAp-derived peaks could be observed in the as-prepared Mg-HAp samples with x values lower than6.0(i.e.at Mg-to-Ca ratios lower than6-to-4).With increasing x,the crystallinity of the apatite phase gradually decreased as manifested by increased broadening and decreased intensity of their XRD peaks.At x=6.0,only an amorphous phase(Mg-containing amorphous calcium phosphate,Mg-ACP) could be synthesized.The gradual decrease of crystal-linity is one of the indications suggesting increasing Mg

203040506070 0

2000

4000

6000

8000

10000

(

)

2Θ (deg)

Fig.1.XRD patterns of as-prepared Mg-HAp powders.Values of the x parameter in the simpli?ed Mg-HAp formula are marked.

W.L.Suchanek et al./Biomaterials25(2004)4647–4657 4650

incorporation in the HAp lattice with increasing x, which will be discussed later.In the pure HAp(x=0.0) and in the Mg-HAp samples with x=0.5–1.0,no other XRD peaks except for those derived from the HAp phase were observed,while peaks indicating the presence of unreacted Mg(OH)2were observed in all as-prepared samples with x=2.0and larger.However, even in the as-prepared Mg-HAp samples with x=0.5–1.0,the presence of small quantities of unreacted Mg(OH)2,undetectable by XRD,was revealed by the thermogravimetric analysis.TG curves of these powders exhibited typical sudden weight loss at about370–400 C which corresponds to decomposition of the Mg(OH)2 phase.In the samples with x=4.0and 5.0,weak NH4MgPO4áH2O-derived peaks were detected.When Ca was totally replaced with Mg in the starting slurry (x=10.0),our experiments produced only the NH4MgPO4áH2O phase instead of HAp or ACP. Generally speaking,our results are in good agreement with most studies on wet synthesis of Mg-HAp indicating limited solubility of Mg in HAp and dif?culty in obtaining a HAp phase with a total Mg substitution (x=10.0).However,we have achieved substantially higher level of Mg substitution in the HAp phase than has ever been described in reliable reports,i.e.up to 28.4wt%(for x=5.0)using our mechanochemical–hydrothermal synthesis vs.up to5wt%in other works on Mg-HAp synthesis from aqueous solutions[4,12–18]. The level of Mg substitution in our Mg-HAp powders will be discussed and characterized in more detail in the next section.

FTIR analysis of the as-prepared Mg-HAp powders was in a good agreement with the results of the XRD analysis.Selected FTIR spectra,shown in Fig.2,show bands typical for HAp in addition to a sharp peak derived from OHàgroups in the unreacted Mg(OH)2 present in the Mg-HAp sample with x=2.0.The FTIR spectra exhibit increased broadening of the PO43à-derived bands indicating increased structural disorder, which could be caused by increased concentration of Mg in the HAp lattice.However,this effect will be discussed in more detail in the next section.

Based upon the presented XRD and FTIR results,the as-prepared powders with x=6.0–10.0were excluded from further characterization and the powders with x values ranging between0.5and5.0were subjected to a puri?cation process,whose results are described in the following section.

3.2.Puri?ed Mg-HAp powders

XRD patterns of the puri?ed Mg-HAp powders are shown in Fig.3.These patterns indicate that the

ammonium citrate treatment was successful in removing all residual Mg(OH)2for x values0.0–5.0. The only detectable crystalline phase in the puri?ed Mg-HAp samples for x=0.0–4.0was HAp.Only in the puri?ed sample at x=5.0was there any indi-cation of a secondary crystalline phase in which 350030002500200015001000500 0

100

150

200

250

(

)

Wavenumber (cm-1)

Fig. 2.FTIR spectra of selected as-prepared Mg-HAp powders. Values of the x parameter in the simpli?ed Mg-HAp formula are marked.

203040506070 0

1000

2000

3000

4000

5000

(

)

2Θ (deg)

Fig.3.XRD patterns of puri?ed Mg-HAp powders.Values of the x parameter in the simpli?ed Mg-HAp formula are marked.

W.L.Suchanek et al./Biomaterials25(2004)4647–46574651

NH 4MgPO 4áH 2O-derived peaks were detected.With increasing x ,the XRD peaks became gradually broader and weaker.This effect could be explained by decreased crystallite size (Table 1)and increased lattice disorder associated with increasing Mg substitution in the HAp lattice,which has been widely reported in the literature [3,4].

FTIR spectra of the puri?ed Mg-HAp powders are shown in Fig.4.These are typical spectra of HAp

showing PO 43à

-derived bands at 478,566,605,963,and 1030–1090cm à1and adsorbed water bands at 1630and 3000–3700cm à1[3,4].Lowintensity of both OH à-derived bands at 630and 3570cm à1,which are clearly visible only in the nominally stoichiometric HAp powder (x =0.0),as well as loss of resolution of the

PO 43à-derived bands with increasing x ,were observed.These effects are typical for Mg-HAp synthesized by wet methods and can be explained by decrease of crystal-linity due to increased Mg substitution in the HAp lattice [4].Very lowintensity of the OH à-derived bands could also be caused by the presence of adsorbed water combined with the high surface area of the Mg-HAp powders at higher x values (Fig.4,Table 1).Another observed feature of the FTIR spectra of the puri?ed Mg-HAp powders was increase of intensity of the band around 870cm à1with increasing x (Fig.4).In this

position,a HPO 42à-derived band overlaps with a CO 32à

-derived band [4].Since the intensity of the major CO 32à-derived bands between 1420and 1480cm à1in the puri?ed Mg-HAp did not increase with increasing x (Fig.4),intensity increase of this band can be attributed

to the increasing HPO 42à

substitution in the HAp lattice.

In HAp prepared by wet methods,increasing HPO 42à

substitution usually accompanies increasing Mg sub-stitution [4],therefore these results serve as another evidence con?rming increasing Mg substitution in the HAp lattice with increasing x values.The presence of

small amounts of CO 3

2à

ions in all puri?ed Mg-HAp powders was due to carrying out all steps of the experimental procedure in air.The position of the

CO 32à-derived bands indicates that CO 32à-for-PO 43àsubstitution dominates in the Mg-HAp powders but some fraction of the OH àgroups might be replaced by

the CO 3

2à

groups,which is usually observed in carbonated HAp powders prepared by wet methods [3,4].The absence of any sharp OH àstretches in the puri?ed Mg-HAp spectra is evidence that the ammo-nium citrate treatment was successful in removing all of the residual Mg(OH)2.

Thermogravimetric analysis of the puri?ed Mg-HAp powders in the temperature range of 25–950 C,which is summarized in Fig.5,con?rmed results of the XRD and FTIR analyses.No weight changes,previously seen in as-prepared Mg-HAp samples and associated with decomposition of Mg(OH)2,were observed in any of the TG curves recorded for the puri?ed Mg-HAp,con?rming complete removal of the residual Mg(OH)2by the ammonium citrate treatment.The total weight loss (D m )increased with increasing x and was 4.18wt%

Table 1

Properties of the puri?ed Mg-HAp powders x in the Mg-HAp formula Speci?c surface area (m 2/g)Equivalent spherical diameter,d BET (nm)Average particle size (nm)Median particle size (nm)0.09121.06856630.59919.23992181.010318.52191022.012914.74563323.011516.592210654.02697.199811995.0

—a

Quantities of powder not suf?cient for characterization.

350030002500200015001000

500

50100150200250300350

400450

T r a n s m i t t a n c e (a r b . u n i t s )

Wavenumber (cm -1

)

Fig.4.FTIR spectra of puri?ed Mg-HAp powders.Values of the x parameter in the simpli?ed Mg-HAp formula are marked.

W.L.Suchanek et al./Biomaterials 25(2004)4647–4657

4652

for x =0.0(i.e.nominally stoichiometric HAp),11.04wt%for x =1.0,and 14.3wt%for x =3.0.In all the powders,signi?cant loss of weight up to approxi-

mately 450 C was very likely due to the loss of adsorbed (up to E 200 C)and lattice water [3].Loss of the carbonate ions could contribute to the total weight loss

above 550 C.HPO 4

2à

groups,which are incorporated into HAp lattice together with Mg transform into pyrophosphate in the temperature range of 400–700 C with a formation of water according to the following

reaction:2HPO 42à-P 2O 74à

+H 2O [3].Above 600–700 C,nonstoichiometric Mg-HAp decomposes,usual-ly with the formation of whitlockite and water [4,34].However,at higher Mg concentrations,formation of Ca 4Mg 5(PO 4)6or b -Ca 2P 2O 7was reported [34].The TG curves shown in Fig.5represent three different kinds of thermal behavior of the puri?ed Mg-HAp powders.The nominally stoichiometric HAp powder (x =0.0)exhib-ited above 450 C minimal weight loss,which is typical for stoichiometric HAp.The Mg-HAp with x =1.0exhibited a sudden weight loss above 700 C which might indicate decomposition into whitlockite.On the other hand,the Mg-HAp powder with x =3.0exhibited a gradual weight loss over the whole temperature range which was probably due to presence of large quantities of lattice water,Mg 2+,and lowcrystallinity [34].

Results of the XRD analysis of the heat-treated Mg-HAp powders (Fig.6)were in a good agreement with the results of the TG analysis.Heat treatment of the as-

prepared Mg-HAp powders at 900 C for 1h in air resulted in narrowing their HAp-derived XRD peaks,which could be attributed to increasing crystal size.Absence of any b -TCP-derived XRD peaks in the heat-treated HAp powder with nominal stoichiometric composition (x =0.0)serves as evidence that the Mg-free HAp prepared by the mechanochemical–hydro-thermal route was stoichiometric.However,with increasing content of Mg substitution,the Mg-HAp phase became thermally unstable.At lowMg concen-trations,i.e.x =0.5–1.0,only weak peaks of whitlockite [(Mg,Ca)3(PO 4)2]were observed in the heat-treated Mg-HAp,in addition to the HAp-derived peaks.Thermal behavior of this kind is typical for Mg-HAp with low levels of Mg substitution and was reported in the literature for both arti?cially prepared and natural (bone,enamel)Mg-HAp samples [4,34].In the heat-treated Mg-HAp powders at higher Mg concentrations,i.e.x =2.0–4.0,no HAp-derived peaks were visible.At x =2.0,the heat-treated Mg-HAp sample comprised of whitlockite as a major phase and Ca 4Mg 5(PO 4)6as a minor phase.With increasing Mg-substitution in the heat-treated powders above x =2.0,in addition to the whitlockite phase more Ca 4Mg 5(PO 4)6was present and b -Ca 2P 2O 7also started to form.At x =4.0,the major phases were b -Ca 2P 2O 7and Ca 4Mg 5(PO 4)6,while whitlockite was the minor phase.Formation of b -Ca 2P 2O 7and Ca 4Mg 5(PO 4)6,both in natural and

I n t e n s i t y (a r b . u n i t s )

2Θ (deg)

Fig.6.XRD patterns of heat-treated Mg-HAp powders (900 C,1h,air).Values of the x parameter in the simpli?ed Mg-HAp formula are marked.

020*******

8001000

-15

-10

-5

W e i g h t c h a n g e (%)

Temperature (o

Fig.5.TG curves of selected puri?ed Mg-HAp powders.Values of the x parameter in the simpli?ed Mg-HAp formula are marked.

W.L.Suchanek et al./Biomaterials 25(2004)4647–4657

4653

arti?cial HAp samples after subjecting them to interac-tions with Mg ions at high Mg concentrations was earlier observed by Baravelli et al.[34].XRD peaks of whitlockite were shifted towards higher 2Y values with respect to Mg-free b -TCP (JCPDS card #09-0169),which serves as an evidence of Mg substitution in the whitlockite phase.It is known that large amounts of Mg can substitute into the whitlockite and stabilize its lattice [4].The XRD analysis of the heat-treated Mg-HAp powders con?rms presence of large amounts of Mg in puri?ed Mg-HAp powders.

XRD,FTIR,and TG analysis indicate that the puri?cation procedure using ammonium citrate is very ef?cient in preferentially removing the Mg(OH)2phase from a mixture with HAp.There is minimal difference in the appearance of the HAp-derived XRD peaks before (Fig.1)and after (Fig.3)the puri?cation step.Similarly,there is minimal difference between the HAp-derived bands in the FTIR spectra acquired from the Mg-HAp powders before (Fig.2)and after (Fig.4)the puri?ca-tion process.These results indicate that this puri?cation process can be accomplished without signi?cantly affecting the Mg-HAp phase.

Chemical analysis of the puri?ed Mg-HAp powders indicates increasing Mg concentration with increasing x from 0.24wt%at x =0.5to 28.2wt%at x =5.0(Fig.7).The relative amounts of Mg in the Mg-HAp were higher than in the starting slurries.A large error of the chemical analysis alone (estimated to be even up to 735%)cannot explain these results.Neither can the presence of NH 4MgPO 4áH 2O detected in the x ?5:0

sample,because based upon the recalculation of chemical analysis results for NH 4(1.8wt%at x =5.0),only up to 2.4wt%of Mg could be associated with the NH 4MgPO 4áH 2O phase.Nevertheless,assuming that all the Mg is incorporated in the HAp lattice and not adsorbed on the surface,under our experimental conditions magnesium seems to be dissolved in HAp equally or even preferentially with respect to calcium.Such conclusion is rather unusual,because the dissolu-tion of Mg in HAp is limited in most cases to a feww t%due to a destabilizing effect of Mg on the HAp lattice [3,4].Achieving such a high level of Mg substitution in HAp might be due to our unique preparation conditions and exceptional capabilities of the mechanochemical–hydrothermal technique,which has not been investi-gated in prior Mg-HAp synthesis research.A comple-mentary chemical analysis for Mg,accomplished using the DC plasma emission spectroscopy technique for puri?ed Mg-HAp samples with x =1.0,2.0,3.0,and 5.0[35]was in a very good agreement (within the measure-ment error)with the XRF and XPS results presented in this paper,as shown in Fig.7.

Position of the Mg in HAp is a subject of much controversy.There are several works,which report presence of Mg on the surface of HAp crystals [36,37].28

Mg exchange from HAp precipitated under physiolo-gical conditions in the presence of Mg 2+showed that 90%of Mg was located on the surface of HAp [38].Conversely,LeGeros [12]and Okazaki [39]report partial substitution of Ca with Mg in the HAp lattice,which greatly affects crystallinity and solubility of HAp.To clarify location of Mg in the Mg-HAp powders,we performed XPS measurements in an attempt to elucidate any Mg enrichment near the surface of the powder.Our results indicated that the concentration of Mg in the puri?ed Mg-HAp powder at x =2.0inside the crystals

(90(A

depth)was 4.5at%(>6wt%),which is larger than 3.9at%of Mg close to the surface (30(A

depth).In the same Mg-HAp powder,however,calcium concen-tration did not change with the sampling depth.This is contrary to previous low-temperature synthesis results shown by earlier investigators where Mg was shown to preferentially adsorb to the surface of apatite with limited lattice substitution [36–38].These measurements can serve as evidence that in the puri?ed Mg-HAp powders,Mg 2+ions were not preferentially located near the surface.However,some of the Mg 2+ions in the near-surface positions could have been leached out during the puri?cation process,resulting in lower Mg concentration near the surface than in the bulk.

Speci?c surface area of the puri?ed Mg-HAp powders ranged between 91m 2/g for x =0.0and 269m 2/g for x =4.0,which corresponds to an estimated equivalent spherical diameter (d BET )of 7–21nm (Table 1).With increasing x ,i.e.with increasing Mg substitution,speci?c surface area increased (Table 1).The particle

Wt.%Mg in the 1015

2530

W t .%M g i n p u r i f i e d M g -H A p

starting slurry (with respect to Mg-HAp)

Fig.7.Results of the chemical analysis for magnesium of puri?ed Mg-HAp powders.Solid circles connected by solid lines represent experimental points obtained by X-ray ?uorescence spectroscopy (this work).Solid triangle represents results obtained by XPS (this work).Open squares connected by doted lines represent experimental points obtained by DC plasma emission spectroscopy (data from Ref.[35]).The dashed line is a line of stoichiometric substitution,i.e.corresponds to the concentration of Mg in the starting slurry being equal to the concentration of Mg in the puri?ed Mg-HAp.Values of the x parameter in the simpli?ed Mg-HAp formula are marked.

W.L.Suchanek et al./Biomaterials 25(2004)4647–4657

4654

size distributions were in all cases single-modal with median values ranging between102and1199nm. Corresponding average particle sizes were in the range of219–998nm(Table1).With increasing x,the measured particle size initially decreased(for x o1.0) and then increased(for x>1.0).These measurements very strongly indicate the presence of aggregates and/or agglomerates consisting of nanosized HAp primary particles with a size of p20nm.

The results of FESEM observations,shown in Fig.8, were in a good agreement with the results of both BET and DLS measurements.The puri?ed Mg-HAp powders contained large agglomerates,1–2m m in diameter, consisting of nanosized Mg-HAp crystals(Figs.8b,c). Both equiaxed(Figs.8a,c)and acicular(Fig.8b) morphologies of the Mg-HAp particles were observed.3.3.Advantages of the mechanochemical–hydrothermal synthesis of Mg-HAp powders

The Mg-HAp powders synthesized in the present work by the mechanochemical–hydrothermal method share several common features.All of them consist of nanosized crystals forming larger aggregates and agglomerates.The concentration of Mg in the HAp lattice could be controlled in a wide range by changing concentration of reactants;however,independent con-trol of the crystal size,morphology,and agglomerate size distribution will require further research.These phenomena seem to be typical for the mechanochem-ical–hydrothermal synthesis of HAp powders and were observed also in our earlier work on pure HAp and carbonated HAp powders[31,32].The absence of any unreacted species in the puri?ed Mg-HAp powders indicates the effectiveness of ammonium citrate treat-ment and demonstrates that a Mg-substituted HAp with elevated and controlled Mg levels can be synthesized entirely at room temperature by this processing route. Most importantly,the mechanochemical–hydrothermal method applied to the Mg-HAp synthesis allows achieving Mg-substitution levels signi?cantly higher than obtained so far by any other synthesis method. The synthesized Mg-HAp powders with controlled Mg-substitution could?nd applications in hard tissue implants as biocompatible ceramics,composite consti-tuents,or coatings,as granular?ll for direct incorpora-tion into tissues,in dental composites,dentifrice,etc.

[40].

The synthesis reactions occur in an alkaline environ-ment,which is very important because high pH values favor formation of an apatitic phase[3,4].The mechan-ochemical–hydrothermal synthesis of Mg-HAp using Ca(OH)2/Mg(OH)2and(NH4)2HPO4occurs with par-ticipation of the liquid phase because one of the reactants,namely(NH4)2HPO4dissolves easily in water. Solubility of Mg(OH)2,Ca(OH)2,and(NH4)2HPO4in water at room temperature is0.009,1.85,and575g/l, respectively[41].It is worth mentioning that both Ca(OH)2and Mg(OH)2are soluble in ammonia salts [41].Our results emphasize the importance of the aqueous solution,which actively participates in the synthesis reaction by dissolving one of the reactants, which is not observed with conventional mechanochem-ical synthesis of HAp[22–29].

Previous research on mechanochemical and mechan-ochemical–hydrothermal synthesis of HAp accom-plished in other groups[22–29]did not take advantage of water as a solvent.The aqueous solution can actively participate in the mechanochemical reaction by accel-eration of dissolution,diffusion,adsorption,reaction rate,and crystallization[42].The mechanochemical activation of slurries can generate local zones of high temperatures(up to450–700 C)and high pressures due

Fig.8.FESEM photographs of the puri?ed Mg-HAp powders

showing agglomeration of nanosized crystals:(a)x=0.0;(b)x=1.0;

and(c)x=3.0.

W.L.Suchanek et al./Biomaterials25(2004)4647–46574655

to friction effects and adiabatic heating of gas bubbles (if present in the slurry),while the overall temperature is close to the room temperature[43].It is worth mentioning that the mechanochemical–hydrothermal route produces from the same volume comparable amounts of HAp powder as the hydrothermal proces-sing,but is easier for scaling-up,since it does not require

a pressurized reactor.

4.Summary

Mg-HAp powders with magnesium concentration between E0.2and E28wt%have been prepared by the mechanochemical–hydrothermal route at room temperature.The powders consisted of both equiaxed and acicular crystals,less than E20nm in diameter, which formed larger aggregates/agglomerates resulting in average particle sizes in the range of219–998nm. Concentration of Mg in the HAp lattice was slightly lower near the surface than in the bulk of the HAp crystals.Our work demonstrates applicability of the mechanochemical–hydrothermal synthesis for reprodu-cible and low-cost synthesis of Mg-HAp powders in large batch-sizes,with Mg substitution levels as high as 28.4wt%.

Acknowledgements

This research was supported by the Johnson& Johnson Center for Biomaterials and Advanced Tech-nologies,the Center for Biomedical Devices at Rutgers University,and the National Institute of Health.The authors are greatly indebted to James Kim for experi-mental assistance,Eric Gulliver and Mohit Jain for obtaining the FESEM images,and to Kelly Brown for the XPS analysis.

References

[1]Aoki H.Science and medical applications of hydroxyapatite.

Tokyo,Japan:Japanese Association of Apatite Science;1991. [2]Suchanek W,Yoshimura M.Processing and properties of

hydroxyapatite-based biomaterials for use as hard tissue replace-ment implants.J Mater Res1998;13:94–117.

[3]Elliott JC.Structure and chemistry of the apatites and other

calcium orthophosphates.Amsterdam:Elsevier;1994.

[4]LeGeros RZ.Calcium phosphates in oral biology and medicine.

Basel,Switzerland:Karger AG;1991.

[5]Kanzaki N,Onuma K,Treboux G,Tsutsumi S,Ito A.Inhibitory

effect of magnesium and zinc on crystallization kinetics of hydroxyapatite(0001)face.J Phys Chem B2000;104:4189–94.

[6]Bon?eld W,Gibson IR.Process for the preparation of magnesium

and carbonate substituted hydroxyapatite.Patent Application No.WO99/32400;1999.

[7]Dolci G,Mongiorgi R,Prati C,Valdre G.Odontostomatologic

use of apatite-based nanostructured materials.Patent Application No.WO00/03747;1998.

[8]Althoff J,Quint P,Krefting ER,H.ohling HJ.Morphological

studies on the epiphyseal growth plate combined with biochemical and X-ray microprobe analysis.Histochemistry1982;74:541–52.

[9]Liu CC,Yeh JK,Aloia JF.Magnesium directly stimulates

osteoblast proliferation.J Bone Mineral Res1988;3:S104. [10]Serre CM,Papillard M,Chavassieux P,Voegel JC,Boivin G.

In?uence of magnesium substitution on a collagen-apatite biomaterial on the production of a calcifying matrix by human osteoblasts.J Biomed Mater Res1998;42:626–33.

[11]Martini LA.Magnesium supplementation and bone turnover.

Nutr Rev1999;57:227–9.

[12]LeGeros RZ.Incorporation of magnesium in synthetic and in

biological apatites.In:Fearnhead RW,Suga S,editors.Tooth enamel IV.Amsterdam:Elsevier;1984.p.32–6.

[13]LeGeros RZ,Daculsi G,Kijkowska R,Kerebel B.The effect of

magnesium on the formation of apatites and whitlockites.In: Itokawa Y,Durlach J,editors.Magnesium in health and disease.

London:John Libbey;1989.pp.11–9.

[14]Bertoni E,Bigi A,Cojazzi G,Gandol?M,Panzavolta S,Roveri

N.Nanocrystals of magnesium and?uoride substituted hydro-xyapatite.J Inorg Biochem1998;72:29–35.

[15]Mayer I,Schlam R,Featherstone JDB.Magnesium-containing

carbonate apatites.J Inorg Biochem1997;66:1–6.

[16]Golden DC,Ming DW.Nutrient-substituted hydroxyapatites:

synthesis and characterization.Soil Sci Soc Am J1999;63:657–64.

[17]Bigi A,Falini G,Foresti E,Gazzano M,Ripamonti A,Roveri N.

Magnesium in?uence on hydroxyapatite crystallization.J Inorg Biochem1993;49:69–78.

[18]Okazaki M,Takahashi J,Kimura https://www.wendangku.net/doc/ba6494679.html,parison of crystal-

lographic properties of Mg,Fe,Na,CO3,F,and Cl-containing apatites.J Osaka Univ Dent Sch1986;26:79–89.

[19]Patel PN.Magnesium calcium hydroxyapatite solid solutions.J

Inorg Nucl Chem1980;42:1129–32.

[20]Chiranjeevirao SV,Hemmerle J,Voegel JC,Frank RM.A

method of preparation and characterization of magnesium-apatites.Inorg Chim Acta1982;67:183–7.

[21]Tepstra RA,Driessens FCM.Magnesium in tooth enamel and

synthetic apatites.Calcif Tissue Int1986;39:348–54.

[22]Toriyama M,Kawamura S.Synthesis of b-tricalcium phosphate

by use of wet milling.J Ceram Soc Jpn1986;94:1004–7.

[23]Toriyama M,Kawamura S.Sinterable powder of mechanochemi-

cally synthetic b-tricalcium phosphate.J Ceram Soc Jpn 1987;95:741–5.

[24]Toriyama M,Kawamura S,Ito Y,Nagae H.Thermal change of

calcium de?cient apatite obtained by mechanochemical treatment.

J Ceram Soc Jpn1989;97:554–8.

[25]Otsuka M,Matsuda Y,Hsu J,Fox JL,Higuchi WI.Mechan-

ochemical synthesis of bioactive material:effect of environmental conditions on the phase transformation of calcium phosphates during grinding.Biomed Mater Eng1994;4:357–62.

[26]Toriyama M,Ravaglioli A,Krajewski A,Galassi C,Roncari E,

Piancastelli A.Slip casting of mechanochemically synthesized hydroxyapatite.J Mater Sci1995;30:3216–21.

[27]Toriyama M,Ravaglioli A,Krajewski A,Celotti G,Piancastelli

A.Synthesis of hydroxyapatite-based powders by mechano-

chemical method and their sintering.J Eur Ceram Soc1996;

16:429–36.

[28]Yokogawa Y,Toriyama M,Kawamoto Y,Suzuki T,Nishizawa

K,Nagata F,Mucalo MR.Preparation of calcium strontium apatite through mechanochemical method.Chem Lett1996;1: 91–2.

[29]Liao J,Ono K,Kanayama G,Isobe T,Senna M.Preparation

of tricalcium phosphate by applying grinding technique and

W.L.Suchanek et al./Biomaterials25(2004)4647–4657 4656

choosing hydroxy-containing initial materials.Chem Sustainable Dev1998;6:233–6.

[30]Avvakumov E,Senna M,Kosova N.Soft mechanochemical

synthesis:a basis for newchemical technologies.Boston,USA: Kluwer Academic Publishers;2001.

[31]Shuk P,Suchanek WL,Hao T,Gulliver E,Riman RE,Senna M,

TenHuisen KS,Janas VF.Mechanochemical–hydrothermal pre-paration of crystalline hydroxyapatite powder at room tempera-ture.J Mater Res2001;16:1231–4.

[32]Suchanek WL,Shuk P,Byrappa K,Riman RE,TenHuisen KS,

Janas VF.Mechanochemical–hydrothermal synthesis of carbo-nated apatite powders at room temperature.Biomaterials 2002;23:699–710.

[33]Hamada K,Senna M.Mechanochemical effects on the properties

of starting mixtures for PbTiO3ceramics by using a novel grinding equipment.J Mater Sci1996;31:1725–8.

[34]Baravelli S,Bigi A,Ripamonti A,Roveri N,Foresti E.Thermal

behavior of bone and synthetic hydroxyapatites submitted to magnesium interaction in aqueous medium.J Inorg Biochem 1984;20:1–12.

[35]Chen CW,Suchanek W,Shuk P,Riman RE,Brown K,

TenHuisen KS,Janas VF.Synthesis and characterization of magnesium-substituted hydroxyapatite.In:Riman RE,editor.

Proceedings of Fifth International Conference on Solvothermal Reactions(ICSTR),East Brunswick,NJ,July22–26,2002.

p.238–41.

[36]Kick CH,Berthke RM,Edington RM,Wilder OHM,Record PR,

Wilder W,Hill TJ,Chase WW.Fluorine in animal nutrition.Ohio Agric Exp St Bull1955;588.

[37]McCann HG.Reaction of?uoride ion with hydroxyapatite.J Biol

Chem1953;201:247–59.

[38]Neuman WF,Mulryan BJ.Synthetic hydroxyapatite crystals IV.

Magnesium incorporation.Calcif Tissue Res1971;7:133–8. [39]Okazaki M,Takahashi J,Kimura H.Unstable behavior of

magnesium-containing hydroxyapatites.Caries Res1986;20: 324–31.

[40]Riman RE,Suchanek W,Shuk P,TenHuisen KS.Magnesium-

substituted https://www.wendangku.net/doc/ba6494679.html, Patent Application https://www.wendangku.net/doc/ba6494679.html,2002/ 0127260A1,2002.

[41]Weast RC,editor.CRC Handbook of chemistry and physics,68th

ed.(1987–1988).Boca Raton,FL:CRC Press;1987.

[42]Riman RE,Suchanek WL,Lencka MM.Hydrothermal crystal-

lization of ceramics.Ann Chim Sci Mater2002;27:15–36. [43]Kosova NV,Khabibullin AKh,Boldyrev VV.Hydrothermal

reactions under mechanochemical treating.Solid State Ionics 1997;101–103:53–8.

W.L.Suchanek et al./Biomaterials25(2004)4647–46574657

2004-01 热重分析很好 水热制备镁取代HA

2004-01 热重分析很好水热制备镁取代HA